Acute Leukemia (Lymphoblastic and Myeloid, ALL and AML) is the most common type of cancer in children (age 0-14 years), but the causes of the disease are uncertain in the vast majority of cases. As leukemia is a cancer of the immune system, its risk is profoundly affected by patterns of infections. Epidemiology studies have shown that exposure to a wide variety and number of childhood contacts (e.g., daycare attendance and older siblings) are protective against childhood ALL. More recent epidemiologic evidence, however, suggests that vigorous response to infectious agents, as measured by infections that require care from medical professionals, is a risk factor for both childhood ALL and AML. Additionally, in our studies we found that ALL cases had decreased levels of the immunosuppressive cytokine IL-10 at birth when compared to control children, suggesting that they enter the world with a congenital immune aberration. In the current proposal, we wish to develop these two critical observations (low IL-10, more severe childhood infections) into a connected and recognizable causal pathway, using both epidemiologic observation in humans, and a manipulable mouse model. In Aim 1, we will measure IL-10 levels in 260 case and 390 control children's archived neonatal dried blood spots (DBS) collected at birth and stored for all California children, and perform association analyses between the IL-10 levels and physician diagnosed infections after birth and before leukemia diagnosis. This aim will be completed using two unique resources: (i) the California Genetic Diseases Branch archived newborn blood resource, which store DBS samples from all California-born children, and (ii), the Kaiser Permanente Northern California medical record database. We will perform this analysis in both children who contracted leukemia and those who did not, to determine whether infections are an obligate intermediary between IL-10 levels and risk of leukemia. In Aim 2, we will use a mouse model that recapitulates the most common subtype of childhood leukemia, those with ETV6-RUNX1 (TEL-AML1) translocations, to study this same causal pathway. ETV6-RUNX1-carrying mice will be mated with IL-10 knockout mice, yielding a new strain that has lower IL-10 at birth along with a pre-leukemic mutation, directly mirroring the human situation. These mice will be challenged with immune stimuli to test the putative cooperative effect of infection with these two pre-leukemic attributes. The animals will be assayed following the challenge to ascertain the interactive effects of the pre-leukemic mutation, IL-10 levels, and immune challenge on the expansion of pre-leukemic immature B-cells and on leukemogenesis. The use of this mouse model will inform and be informed by epidemiologic observations, permitting the construction of a synergistic knowledge base on a new leukemia causal hypothesis within the proposed five year grant period.